Skip to main content
. 2021 May 15;21(10):3444. doi: 10.3390/s21103444

Table 2.

Characteristics of included studies.

First Author [Ref] Aim Participant Characteristics Wearable Device Sensors Type (Short) Experimental Intervention Control Intervention Timing Balance and Gait Outcome Measure Evaluation Time Points
Azerpaikan et al. [40] To study the effect of a neurofeedback training on balance problems associated with Parkinson’s disease. TOT n = 16 patients with PD
EG: n = 8
Mean age: 74.23 ± 3.51 years
Female: n = 4
CG: n = 8
Mean age: 75.16 ± 3.64 years
Female: n = 4
Biograph Infiniti Software system (version 5.0), the ProComp differential amplifier (Thought Technology Ltd, Montreal, Quebec) for NeuroFeedback training sessions (FlexComp Infiniti encoder, TT-USB interface unit, fiber optic cable, USB cable) EEG SENSORS EG: Neurofeedback training with EEG generator EG: Sham Neuro Feedback Training using sham EEG generator D:8 sessions
T: 30 min
F: 3 days/week
BBS*, Limit of stability * PRE, POST
Byl et al. [41] To evaluate the effectiveness of supervised gait training with and without visual kinematic feedback TOT: n = 24 stroke and PD
EG: n = 12
Stroke: n = 5
Mean age: 66.2 ± 5.0 years
Female: n = 3
PD: n = 7
Mean age: 68.5 ± 3.6 years
Female: n = 4
CG: n = 12
Stroke: n = 7
Mean age: 60.8 ± 5.4 years
Female: n = 5PD: n = 5
Mean age: 70.0 ± 2.9 years
Female: n = 2
Wireless joint angle sensors and Smart Shoes IMU, PRESSURE SENSORS EG: visual kinematic feedback on the computer screen during progressive and task-oriented balance and gait training activities CG: balance and gait training activities D: 12 sessions (6–8 weeks)
T: 90 min (30 min visual kinematic feedback in the EG)
F: NA
Gait Speed; Step Length; TBS, 6MWT, DGI, 5TSS, TUG, BBS, FOG-Q PRE, POST
Carpinella et al. [42] To test the feasibility of a wearable biofeedback system in a typical rehabilitation gym and analyze the effect on balance and gait outcome measures compared to physiotherapy without feedback. TOT: n = 42 subjects with PD
EG: n = 17
Mean Age: 73 ± 7.1 years
Female: n = 3
CG: n = 20
Mean Age: 75.6 ± 8.2 years
Female: n = 11
Gamepad System (IMU, PC and Customized software) IMU SENSORS EG: balance and gait functional tailored exercises using Gamepad System CG: personalized balance and gait exercises defined by the clinical staff D: 20sessions
T:45 min
F:3 times/week
BBS *, Gait Speed, UPDRSIII, TUG, ABC, FOGQ, COP ML sway *, COP AP sway, Tele-healthcare Satisfaction Questionnaire-Wearable Technologies PRE, POST, FU
Cha et al. [46] To compare the effectiveness of auditory feedback stimulation from the heel and forefoot areas in terms of ambulatory functional improvements in stroke patients TOT n = 31 stroke subjects EG1: n = 11 Mean age: 64.6 ± 10.6 Female: n = 3 EG2: n = 10 Mean age: 63.0 ± 4.7 Female: n = 3 CG: n = 10 Mean age: 61.8 ± 9.8 Female: n = 4 A PedAlert Monitor 120 (PattersonMedical Holdings Inc.,
Warrenville, IL, USA)
PRESSURE SENSORS EG1: gait training with active weight bearing on the paretic heel with auditory feedback EG2: gait training with auditory feedback from paretic metatarsals CG= gait intervention D: 6 weeks T: 50 min (30 min of conventional therapy + 20 min of gait intervention with or without auditory feedback) F: 3 × week Gait speed, FGA *, TUG, COP length EO *, COP length EC, COP velocity EO *, COP velocity EC PRE, POST
Cho et al. [47] To examine whether visual biofeedback tracking training can improve gait performance in chronic stroke patients. TOT n = 10 Stroke subjects
EG: n = 5
Mean age: 46.2 ± 7.3 years
Female: n = 2
CG: n = 5
Mean age: 48.8 ± 6.3 years
Female: n = 1
A double-axis electrogoniometer (Biometrics Ltd. Ladysmith, VA) was used to record the instant degrees
of knee joint flexion–extension. Series of PC generated sine waves at 0.2 Hz were displayed on a PC monitor at 80 cm distance from the
eyes of the subject
ELECTRO-GONIOMETER EG: visual biofeedback tracking training CG: not Reported D:20 sessions (4 week) T: 39 min F: 5 days/week Motoricity Index, Modified Motor Assessment Scale, Gait Speed. PRE, POST
Choi et al. [48] To compare gait intervention with auditory feedback induced by active weight bearing on the paralyzed side with the effects of the general gait training method TOT n = 24 stroke subjects EG: n = 12 Mean age: 62.8 ± 4.8 Female: n = 4 CG: n = 12 Mean age: 59.7 ± 10.2 Female: n = 4 PedAlert Monitor 120, (Patterson Medical Holdings, Inc.). PRESSURE SENSORS EG: gait intervention with auditory feedback CG: general gait training over the ground. D: 6 weeks T: 50 min (30 min of conventional therapy + 20 mins of gait training with or without bfb) F: 3 × week Gait Speed (10MWT) *, FGA *, TUG *, COP length EO *, COP length EC* PRE, POST
Cozean et al. [49] To study the effect of EMG Biofeedback and FES as therapies for gait dysfunction in patients with hemiplegia after stroke. TOT n = 36
Patients with stroke EG1: n = 9
Mean age:51 ± (not specified)
Female: n = 4
EG2: n = 10 Mean age: 52 ± (not specified)
Female: n = 2 G3: n = 8
Mean age: 56 ± (not specified)
Female: n = 6 CG: n = 9
Mean age: 62 ± (not specified)
Female: n = 2
Not reported EMG SENSORS EG1: EMG biofeedback during static and dynamic activities EG2: FES during static and dynamic activities EG3: EMG biofeedback+FES during static and dynamic activities CG: conventional Physical Therapy D: 6 week T: 30 min F: 3 × Week ankle angle (swing phase) *, knee angle (swing phase) *, stride length, gait speed* PRE, POST
El-Tamawy et al. [43] To examine the influence of paired proprioceptive cues on gait parameters of individuals with PD. TOT n = 30 subjects with PD EG: n = 15 Mean age: 61.4 ± 7.28 Female: not specified CG: n = 15 Mean age: 63.2 ± 5.6 Female: not specified The vibratory device, OPTEC Co. LLtd. PRESSURE SENSORS EG: individually designed physiotherapy and traditional gait training plus treadmill training with vibratory stimuli CG: individually designed physiotherapy and traditional gait training including instructions to walk with long steps. EG: D: 8 weeks T: 51–70 min F: 3 sess/week, CG: D: 8 weeks T: 45 min F: 3 sess/week Cadence *, Stride length *, Gait speed *, Walking distance * PRE, POST
Ginis et al. [44] To test the feasibility of CuPID system in the home environment and verify differential effects of CuPID training versus conventional home-based gait intervention TOT n = 40 subjects with PD EG: n = 22 Mean age: not specified
Female: not specified
CG: n = 18
Mean age: not specified
Female: not specified
The CuPiD system consisted of a smartphone (Galaxy S3-mini,
Samsung, South Korea), a docking station and two IMUs (EXLs3,
EXEL srl., Italy)
IMU SENSORS EG: received weekly home visit and patients were instructed to walk with the CUPID system CG: received weekly home visit by the researcher who gave advice on gait and freezing, and patients were instructed to walk without using the CUPID system D: 6 weeks T: 30 min F:3 times/week MiniBEST *, FSST, FES-I, 2MWT, UPDRS III, NFOG-Q, Comfortable gait and Dual task activities (gait speed, stride length, DS time); PRE, POST, FU
Intiso et al. [50] to evaluate the efficacy of electromyographic biofeedback compared with physical therapy. TOT n = 16
Patients with stroke
EG: n = 8
Mean age: 61.3 ± 12.3 years
Female: n = 4CG: n = 8
Mean age: 53.5 ± 18.5 years
Female: n = 3
table Satem PT
1015 and a walking Satem EMG Combitrainer PT 9115.
EMG SENSORS EG: EMG Biofeedback and Physical Therapy (standard exercise bobath, facilitation, and inhibition techniques, neurofacilitatory techniques) CG: Physical therapy (standard exercise bobath, facilitation and inhibition techniques, neurofacilitatory techniques) D: 2 months T: 60 min F: daily physical therapy (Only EG 30 session of EMG BFB) Basmajian scale *, Gait speed, step length, ankle angle (swing phase) *, ankle angle (heel contact) PRE, POST
Jonsdottir et al. [51] to assess the efficacy of EMG-BFB applied in a task-oriented approach based on principles of motor learning to increase peak ankle power of the affected leg and gait velocity in patients with chronic mild to moderate hemiparesis TOT n = 20
Patients with stroke EG: n = 10
Mean age: 61.6 ± 13.1 years
Female: not specified
CG: n = 10
Mean age: 62.6 ± 9.5 years
Female: not specified
BFB device: (SATEM Mygotron, SATEM srl, Rome, Italy); EMG, system (band-pass filtered at 20 to 950 Hz and then amplified with a gain of 40 000 (50 mVrms range), EMG SENSORS EG=Task-oriented gait training with EMG BFB device CG= conventional physical therapy (at least 15 mins of gait training in each session) D: 20 session T: 45 min F: 3 × Week Gait speed *, ankle power peak at push-off *, stride length *, knee flexion peak PRE, POST, FU
Jung et al. [52] to examine the effect of gait training using a cane with an augmented pressure sensor to improve weight bearing on the nonparetic leg in patients with stroke TOT 22 stroke subjects EG: n = 12 Mean age: 56.4 ± 11.1 Female: n = 4 CG: n = 10 Mean age: 56.3 ± 17.1 Female: n = 3 An instrumented cane, outfitted with a pressure sensor (CD 210-K200, Dacell Co. Ltd., Korea) connected to an indicator (DN30W, Dacell Co. Ltd., Korea) PRESSURE SENSORS EG: gait training with auditory feedback CG: gait training without auditory feedback D: 4 weeks T: 60 min (30 min +30 min gait training with or without bfb) F: 5 × week Peak force cane *, Gait speed *, single support time * PRE, POST
Ki et al. [53] to examine the effects of auditory feedback during gait on the weight bearing of patients with hemiplegia resulting from a stroke. TOT n = 30 stroke subjects EG: n = 12 Mean age: 55.3 ± 9.2 Female: n = 4 CG: n = 13 Mean age: 60.1 ± 12.3 Female: n = 2 A pressure gauge Ped-AlertTM120 (ORBITEC, USA) PRESSURE SENSORS EG: neurodevelopmental treatment with auditory feedback CG: neurodevelopmental treatment D: 4 weeks T: NA F: NA. TUG *, Stance phase duration, Single support time. PRE, POST
Lee et al [54] to examine the effect of neurofeedack training on brain waves control and gait performed under dual-task conditions. TOT n = 20 stroke subjects EG: n = 10 Mean age: 53.2 ± 6.46 Female: n = 4 CG: n = 10 Mean age: 54.7 ± 3.77 Female: n = 3 The Procomp Infiniti system (SA7951 version 5.1, Thought Technology, Canada) was used for neurofeedback training. The QEEG-8 (LXE3208, LAXHA Inc., Korea) system was used to measure brain waves EEG SENSORS EG: neurofeedback training CG: pseudo-neurofeedback training (sham neurofeedback) D: 8 weeks T: 30 min F: 3 × week gait speed *, cadence *, stance phase percentage, and plantar foot pressure (dual task) * PRE, POST
Lupo et al. [55] To evaluate the efficacy of training involving the use of a combined biofeedback system versus conventional balance training TOT n = 15 stroke subjects EG: n = 9 Mean Age: 52.56 ± 13.92 Female: n = 3 CG: n = 6 Mean Age: 65.66 ± 9.64 Female: n = 1 The RIABLO™ (CoRehab, Trento, Italy) system comprised of several inertial measurement unitsand a force platform connected wirelessly to a computer IMU SENSORS AND FORCE PLATFORM EG: balance training with RIABLO biofeedback system using a video interface. CG: conventional balance training without the use of the RIABLO biofeedback system D: 10 sessions T: 20 min F:3 times/week BBS *, RMI, COP length EO *, COP length EC * PRE, POST, FU
Mandel et al. [56] to investigate the efficacy of electromyographic (EMG) versus a novel biofeedback (BFB) approach to improve ankle control and functional gait in stroke patients TOT n = 37 stroke subjects
EG1: n = 13
Mean age: 54.7 ± 13.9 years
Female: n = 5 EG2: n = 13 Mean age: 57.5 ± 14.2 years
Female: n = 5
CG: n = 11 Mean age: 56.8 ± 12.8 years
Female: n = 1
Two channels of EM-BFB, a Lamoureux-type parallelogram electrogoniometer was and a computerized
system to provide audiovisual feedback of ankle position during dorsiflexion and plantar
flexion.
EMG SENSORS and ELECTRO-GONIOMETER EG1: EMG biofeedback training (computer-generated auditory and visual feedback of calf and pretibial muscle activity during active ankle movements) EG2: EMG biofeedback followed by rhythmic positional biofeedback (computer-generated single-channel feedback of dorsiflexion and plantar flexion) CG: No training D: 24 session (EG2 performed 12 session of EMG biofeedback and 12 session of Rythmical positional biofeedback) T: not specified F: 3 × Week Gait speed *, ROM * PRE, POST, FU
Schwenk et al. [58] To evaluate the feasibility and experience in using the new sensor-based training in a sample of patients with clinically confirmed amnestic MCI TOT n = 32 subjects with MCI EG: mean age 77.8 ± 6.9 Female: n = 7 CG: mean age 79 ± 10.4 Female: n = 5 The technology consisted of a 24 inch computer screen, an interactive virtual user interface, and 5 inertial sensors (LegSysTM, BioSensics LLC, MA, USA) IMU SENSORS AND FORCE PLATFORM EG: postural balance exercises during standing (ankle point-to-point reaching tasks and virtual obstacles crossing tasks) using biofeedback training CG: No training EG: D: 4 weeks T: 45 mins F: 2 sessions/week CoM area EO *, CoM area EC, CoM sway ML EO *, CoM swayML EC, CoM sway AP EO *, CoM sway AP EC, Gait Speed, Gait stride time variability, FES-I * PRE, POST
Sungkarat et al. [57] To determine whether improved symmetrical weight bearing somatosensory feedback would result in improved gait and balance in people with stroke TOT n = 35 people with stroke EG: n = 17 Mean Age: 52.12 ± 7.17 years Female: n = 5 CG: n = 18 Mean Age: 53.83 ± 11.18 years Female: n = 6 Insole Shoe Wedge and Sensors (I-ShoWS) PRESSURE SENSORS EG: conventional rehabilitation and gait training with I-ShoWS set-up CG: conventional rehabilitation and gait training without I-ShoWS set-up D:15 sessions T: 60 min/session, (30 min gait training and 30 min conventional) F: 5 days/week Gait speed *, Step Length asymmetry ratio*, Single Support time asymmetry ratio *, Load on paretic leg during stance (% Body Weight) *,
BBS *, TUG *
PRE, POST
van den Heuvel et al. [45] to investigate the feasibility of visual feedback-based balance training (VFT) and to compare the effects of the training program with conventional training TOT n = 33 subjects with PD EG: mean age 63.9 ± 6.39 Female: n = 5 CG:mean age 68.8 ± 9.68 Female: n = 8 Flat-panel LCD monitor connected to a PC (Motek Medical, Amsterdam, The Netherlands), Force plate (Forcelink, Culemborg, and The Netherlands) and Inertial sensors (X sens, Enschede, The Netherlands) IMU SENSORS EG: interactive balance games with explicit augmented visual feedback CG: conventional balance training recommended by the guidelines for physical therapy. D: 5 weeks T: 60 min (45 min balance workstation) F: 2 sess/week BBS; Single leg stance test; Gait speed, FES-I, UPDRSIII* PRE, POST, FU

* = p ≤ 0.005; 5TSS = 5 Time sit to stand; 2MWT = 2 min Walk Test; 6MWT = 6 min Walk Test; AP = Antero-posterior; BFB = Biofeedback; BBS = Berg balance scale; CG = Control group; COM = Centre of movement; COP = Centre of pressure; D = Duration; DGI = Dynamic Gait Index; EEG = Electroencephalogram; EG = Experimental group; EMG = Electromyography, F = Frequency; FES-I = Fall efficacy scale; FGA = Functional gait assessment; FOG-Q = Freezing of gait questionnaire; FSST = Four Square step test; IMU = Inertial movement unit; MCI = Mild cognitive impairment; ML = Medio-lateral; NFOG-Q = New freezing of gait questionnaire; PD = Parkinson disease; RMI = Rivermead mobility index; ROM = Range of motion; T = time; TBS = Tinetti balance scale; TOT = Total; TUG= Timed up and go; UPDRS = Unified Parkinson disease rating scale.